Ocean Sci. Discuss., 8, C731–C742, 2011
www.ocean-sci-discuss.net/8/C731/2011/
© Author(s) 2011. This work is distributed under
the Creative Commons Attribute 3.0 License.
Ocean Science
Discussions
OSD
8, C731–C742, 2011
Interactive
Comment
Interactive comment on “On the export of dense
water from the Weddell and Ross Seas” by R. Kerr
et al.
R. Kerr et al.
[email protected]
Received and published: 19 October 2011
We thank reviewer #2 for suggestions and very constructive comments, which have
significantly contributed to improve this m/s. Our reply following the list of the reviewer
comments.
Referee 2:
Full Screen / Esc
Printer-friendly Version
General comments: The authors describe seasonal and interannual variabilities of water mass export from the Weddell and Ross Seas, based on a state-of-the-art iceocean coupled model (OCCAM). For the interannual variability, they show spatial and
time varying characteristics of dense water export from the basins and shelf water
mass, and find a coherent variation between the two, mainly due to the change of the
C731
Interactive Discussion
Discussion Paper
wind driven ocean circulation. I think the topic is generally interesting and important
for Antarctic research. The structure and analysis method of the paper are totally reasonable and all figures are fine. Therefore, I think this paper is basically worth for
publishing in Ocean Science.
However, I have one critical concern about the definition for estimating (dense water)
export, and my decision for the current manuscript is major revision. In this paper, water export are estimated by calculating the transport across some specified sections. I
think that this calculation leads to underestimate or overestimate for the exports. (For
example, some changes of the position of density front could easily and largely modify
the estimated value.) In particular, I think that the export from dense shelf water formation area should be defined by all export from shallower region to deeper region in
continental shelf (i.e. export across shelf break) to take account of all dense shelf water
mass formation over the continental shelf. Moreover, for the current AABW estimation,
I could not separate the transport over continental shelf (upper layer) or over continental slope and rise (lower layer). I think export in the lower layer should be treated as
AABW transport.
AR: We agree with the reviewer in this point that the m/s does not fully explore the
issue of dense water export. However, it is, in general, out of the scope of this study
to investigate the dense shelf water outflow into the deep ocean using OCCAM model.
These results are presented in a more comprehensive way in R. Kerr’s PhD thesis.
In fact, the model did not reproduce AABW formation very well around the Antarctic
margins, which was expected due to the model spatial resolution, lack of tides, lack of
ice shelves, and weaknesses in the sea ice model. Therefore, we think that deep ocean
convection may also be acting as AABW sources in this particular run. Hence, we
present some results in the m/s showing the dense water outflow into the deep ocean
through the analysis of NW Weddell Sea formation section, which is the only region in
the OCCAM model that displays a satisfactory result related to AABW production near
the shelf-break.
C732
OSD
8, C731–C742, 2011
Interactive
Comment
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
To avoid confusion when we are talking about the AABW flow around the continental
slope (and consequently export to global ocean) and dense shelf water spilling off the
shelf, we have renamed the sections following the reviewer’s 2 suggestion. Thus, we
have now: (i) Weddell and Ross Sea Export Sections (the same as presented in OSD
m/s), (ii) Weddell and Ross Sea Southern Sections (referring to the old "formation"
sections in the southern regions), and (iii) Weddell Sea Formation Section (that refers
to the old NW Weddell Formation Section). We think that re-labeling the sections as
indicated above clarifies that the volume transport in the Export and Southern sections
are primarily the lateral transports of recirculating and new AABW varieties in the Weddell and Ross Seas (e.g. WSBW and WSDW in the Weddell Sea). Thus, we would not
compare the transport across the Southern Sections with estimates of rates of AABW
production in the real ocean, which was a concern pointed out by the reviewer.
OSD
8, C731–C742, 2011
Interactive
Comment
Specific Comments: [1.
Introduction] I think this introduction is well written.
(P1660/L27-28) "..., therefore modelling results ... an alternative way to ...on AABW
export." I think the expression is too strong. This sentence is only valid when the
model shows perfect performance. I recommend that authors rephrase it.
AR: The sentence was rephrased to "Therefore modelling results are an alternative way
to fill those gaps and, in this study, to address the impact of source water variability on
AABW export.
[2. Model description and forcing] (P1661/L11-14) Remove the sentences. Why do you
describe a different model (NCAR-CCSM) performance in section 2?
Full Screen / Esc
AR: The sentence was removed.
Printer-friendly Version
(P1661/L24) "explicitly" Does the effect of ice shelves, icebergs, and runoff from ice
sheet include implicitly in the model?
Interactive Discussion
AR: We rephrased the sentence to clarify this point. Please see below.
“However, cryosphere processes related to ice shelves, icebergs and ice sheet runoff
C733
Discussion Paper
are not included in the model.”
(P1661/L25) "66 levels" Generally, a Z-coordinate model has difficulty to simulate
downslope flow of dense water over continental slope (Wang et. al 2008). Does the
model have enough vertical resolution for reproducing AABW formation processes?
Reference: Q. Wang, S. Danilov, and J. Schroter (2008), Comparison of overflow simulations on different vertical grids using the Finite Element Ocean circulation Model.
Ocean Modelling 20, 313-335
OSD
8, C731–C742, 2011
Interactive
Comment
AR: We agree with the reviewer’s comment and as also highlighted by Reviewer 1 we
include a sentence that talk about some deficiencies of the model used in this study to
better reproduce AABW formation in the shelf-break zone. Please see below the full
paragraph inserted in the m/s.
"It is well known that the shape of the ocean floor and, specially, the seabed corrugations act to steer the main downslope flow of dense shelf waters towards the deep
ocean, thus impacting directly the contribution of AABW. This was recently highlighted
by Muench et al. (2009) for the Northwestern Ross Sea. They point out that corrugations enhance entrainment and reduce along-slope speed of the dense outflow. Thus,
the authors show the importance of considering small-scale local topography when
modeling dense outflows. The bathymetry used in OCCAM does not benefit from recent high resolution swath bathymetric surveys around the Antarctic and therefore may
be locally incorrect and/or too smooth. However, comparing a latitudinal section of the
OCCAM’s bottom volume transport spilling off the shelf (not shown) in the Northwestern Weddell Sea with results obtained from Gordon and Muench (1995; see their Fig.
8) in an approximately similar area, the main direction of OCCAM’s volume transport
in this area agrees with that shown by Gordon and Muench. Additionally, export of
deep and bottom water from the Ross Sea is particularly sensitive to tides (Padman
et al., 2009), however the model output used here does not allow investigation of this
influence. By the way, it is indicated that a Z-coordinate model has some difficulty to
simulate downslope flow of dense water over continental slope (Wang et al., 2008)."
C734
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
(P1662/L9-11) "Sea ice ... in each affected grid cell" Is this description for an initial
condition for sea ice field?
OSD
AR: Yes. This topic was clarified in the m/s as indicated above when we are answering
to Reviewer 1.
8, C731–C742, 2011
[3. Southern Ocean representation] As mentioned above, please consider to redefine
the estimation of AABW and shelf water exports. Since this paper is model-based
study, the more precise and consistent definition for the exports in the model should be
used.
Interactive
Comment
AR: Please see the first answer to Reviewer 2, when we explain the changes made.
[3.1 Modelled hydrographic sections] (P1663/2nd paragraph and Fig. 2) In my reading,
HSSW is also missing in Figs. 2a, 2c, and 2e. Therefore, I could not reach the conclusion in L23-L24 (", the water masses ... represented reasonably well in the sections
(Fig. 2)"
AR: We have changed the sentence (please see below) and include some additional
information that was not present previously in the m/s, but allow us to consider a good
representation of HSSW in the model. “The reasons of HSSW absence in some of
theïĂăïAś/S
˛
diagrams shown (e.g. Fig. 2a) are related to the fact that only a small part
of the export and southern sections are over the continental shelf. Model-observation
water mass properties statistical comparison through Taylor diagrams (R. Kerr PhD
thesis 2010, not shown) also support the good representation of the modelled hydrographical sections used in this study. That is particularly true for those regions with
repeat observations (e.g. Weddell Sea export section).”
Full Screen / Esc
Printer-friendly Version
(P1664/L1-4) "We conclude ..., which provides a high degree of confidence in the analysis of the model time series" There isn’t any result of time series of water masses.
Interactive Discussion
AR: We rephrased the sentence. See below. "We show that OCCAM preserves the
observed general ïAś/S
˛
shape but it also represents the pattern of water mass structure
Discussion Paper
C735
for both coastal and deep ocean in the Antarctic continental margins (Fig. 2), which
provides a high degree of confidence to perform subsequent analysis of the model time
series."
8, C731–C742, 2011
[3.2 AABW source layers depth] For discussing water masses which are separated by
neutral density surfaces, some more information (e.g. contours of 28.00 and 28.27 of
neutral density) should be overlaid in Fig. 3. In alternative way (and its direct way),
authors can discuss water masses averaged over the neutral density layer.
Interactive
Comment
OSD
AR: This section was included in the m/s only to report how the model is representing
the AABW source layers, we have added the neutral density surface isopycnals in Fig.
3 as suggested by the reviewer.
(P1664/L26-27) "the highest salinity plume .." Is there any result or reference that supports this sentence?
AR: We have added the reference below that discuss about the plumes over the western Ross Sea continental slope. Gordon, A. L., Zambianchi, E., Orsi, A., Visbeck,
M., Giulivi, C. F., Whitworth III, T., and Spezie, G. (2004). Energetic plumes over
the western Ross Sea continental slope, Geophysical Research Letters, 31, L21302,
doi:10.1029/2004GL020785.
[4. Cross-section volume transport] [4.1 Full-depth volume transport] The volume
transport of the gyre should be described by one value (e.g. maximum value of
barotropic stream function). I think that the difference between export and formation
sections indicates missing flow branches of the gyre in the estimation.
Full Screen / Esc
Printer-friendly Version
AR: We have added this information, please see below. “The gyre transports at the two
locations are not the same because the sections as shown do not include the whole of
the Weddell/Ross Gyre transports, which can be indicating missing flow branches of
the gyre in that estimation.”
(P1666/L6-9) "Nevertheless, the OCCAM model ..." Judging from the mean, the variC736
Interactive Discussion
Discussion Paper
ance, and the trend (Fig. 4 a-b), the sentence could not be a reason for underestimation
of full-depth volume transport.
AR: The OCCAM values show transports of a similar magnitude of the ADELIE summer measurements only during winter months. So, we include this information in the
paragraph to clarify the sentence. Please see below. "Nevertheless, although OCCAM
simulation show winter transports (Fig. 4a) of a similar magnitude of that reported by
ADELIE summer observation, the OCCAM model transports represents an average
over 17-year period and one might expect them to be less than some observed values
from snapshot cruises or short time series (< 3 years), which are subject to short-term
variability and/or long-term trends."
OSD
8, C731–C742, 2011
Interactive
Comment
[4.2 Bottom layer definition] (P1666/L19-20) "Moreover, OCCAM model.." Add some
references that show the model performance for reproducing bottom waters around
Antarctica.
AR: We have added the reference of Renner et al. (2009).
(P1666/L20-22) "This can be clearly seen ..." A figure based on observation (temperature and density) is needed for the comparison against the model result (Fig. 5).
AR: We have cited the reference of the Southern Ocean Atlas (below) for comparison
with the figure. However, considering the benefit to the whole m/s, we do not think
that an observed climatology Figure is essential in this case. Orsi., A. H., T. Whitworth
III, Hydrographic Atlas of the World Ocean Circulation Experiment (WOCE). Volume 1:
Southern Ocean (eds. M. Sparrow, P. Chapman and J. Gould), International WOCE
Project Office, Southampton, U.K., ISBN 0-904175-49-9. 2004.
[4.3 bottom layer volume transport] (P1667/2nd paragraph and Fig. 7) Instead of performing normalization, I recommend that authors use the right vertical axis (different
scale from the left axis) to show the counterpart timeseries. I think that the difference
of absolute value is important information when comparing the model result with obC737
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
servation.
AR: Done. Figure 7 now show the absolute values instead of normalized ones.
(P1668/L1) "slightly" It seems that the modeled mean transport of 0.5 Sv is much lower
than the observed one (1.95 Sv). It is not a "slightly" difference.
AR: The word was removed.
(P1669/L15-16) Why does the variability export from the Ross Sea formation section
show low?
OSD
8, C731–C742, 2011
Interactive
Comment
AR: As previously reported the model has some deficiencies to reproduce the AABW
production in the shelf-break zones. This is particularly true for the Ross Sea, where
the model AABW production is generally below 0.5 Sv. Thus, this could be reflecting in
low energy associated with AABW export variability in this region. On the other hand,
although the model transports was validated with observations in the Weddell Sea, we
cannot find AABW export time series available for comparison of the AABW export in
the Ross Sea. We assume in this case that there is no reason to consider that the
AABW export variability is not represented, as good as presented in the Weddell Sea,
in other areas around the continent. We added the comments in the m/s. Please, see
below.
"This low energy can be related to the model deficiencies to reproduce the AABW production in the shelf-break zones, probably due to resolution issues. This is particularly
true for the Ross Sea, where the model AABW production is generally below 0.5 Sv.
On the other hand, although the simulated AABW transport was validated with observations in the Northwestern Weddell Sea, we cannot find available time series for
comparison of the AABW export in the Ross Sea"
(P1669/L16-22) Although a reference (Venegas and Drinkwater 2001) is cited for the
explanation of the export variability shown in Fig. 9, I think that some results are
needed to show dynamical link between dense water export and surface variables
C738
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
(wind, temperature, sea ice production, and so on).
AR: As explained above and following both the reviewer’s suggestions we inserted a
sentence in the m/s that talk about lagged correlations between the wind and AABW
export time series in the Weddell Sea. Please see below.
"To show dynamical link between AABW export and surface variables we performed
a lagged correlation between zonal wind stress (defined here as the annual mean
between 60◦ S and 65◦ S - 040◦ W and 020◦ E) and AABW export time series in the
Weddell Sea export section. The result (not shown) unveils the highest correlations
with a five months lag (36%) and around lag-0 (25-35%). The former was the same
period of influence previously reported by Jullion et al. (2010) study, which revealed
that the variability in the wind stress field over the Weddell Sea leads changes in AABW
properties in the Eastern Drake Passage by approximately five months. Moreover, the
same study suggested that the variability in AABW properties over that same area
could be controlled by AABW export from the Weddell Sea."
OSD
8, C731–C742, 2011
Interactive
Comment
[5. Correlations between AABW sources properties and deep ocean export] In my
reading, judging from timeseries in Fig. 10, the high correlation in the section 5 is
achieved by decadal variability (a low state in 1990-1994, and high states in 19881989 and 1999-2005). Is the discussion/timescale of correlation analysis in this section
consistent with those in the previous sections?
AR: Yes, it is. The previous analysis, based on wavelet power spectrum, shows high
energy during periods of 1 and 2-4 yr, but also it is present high variability of a period
of 8 yr (that is not shown, as it lays out of the cone of influence). The total span of
OCCAM time series used (i.e. 17 yr) does not allow us to see and to discuss more
about decadal changes in AABW volume transport.
Printer-friendly Version
(P1671/L4-5) "faster or slower phase of the wind-driven Weddell Gyre" What does this
phrase mean?
Discussion Paper
C739
Full Screen / Esc
Interactive Discussion
AR: We clarified this sentence in the text, please see below. “. . .probably because both
are associated with faster or slower periods of the wind-driven Weddell Gyre.”
(P1671/L16-18) "Finally, another positive feedback ..." I didn’t understand the sentence
at all. If it is important, please add more explanation.
AR: To avoid confusion the sentence was removed.
(P1672/L11-15) "Thus, the correlation maps are" In Hellmer et al 2011, there is a freshening trend (0.09 psu / 17 years) over the northwestern Weddell Sea continental shelf.
In timeseries of salinity anomaly in Fig. 10, however, there is a significant increasing
trend in the salinity over the northwestern Weddell Sea continental shelf. Is it possible
to discuss the phenomena associated with observed freshening trend?
AR: Ours findings reveal a phase correlation between AABW volume export and shelf
waters salinity in the NW Weddell Sea time series, thus evidence that freshening trends
could be used as a proxy to determine variations (decreasing in this case) of AABW
export. We are not interested at this time to determine properties trends from the model
outputs, but to relate possible properties changes and variability with AABW export. In
addition, Hellmer et al. (2011) show the steepest freshening trend between 1989 and
1997, a period that agrees with the highest OCCAM salinity anomaly found between
1989 and 1994 in the model. The observed freshening trend between 1997 and 2006
is not so intense as in the previous years, and the model show alternating periods of
freshening also not so intense after 1997. We added this comment in the m/s. Please,
see below.
"Additionally, Hellmer et al. (2011) show the steepest freshening trend in the tip of
the Antarctic Peninsula between 1989 and 1997, a period that agrees with the highest
OCCAM salinity anomaly found between 1989 and 1994 (Fig. 10). The observed
freshening trend between 1997 and 2006 is not so intense as in the previous years,
and the model show alternating periods of freshening also not so intense after 1997
(Fig. 10)."
C740
OSD
8, C731–C742, 2011
Interactive
Comment
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
(P1672/L21-23) "In general, bottom water export ...’ Add some references.
AR: As we are talking about our results we include the reference of figure 11 to be
checked. (P1673/L13-15) "We have shown that ..." What kind of change of wind are
observed in the forcing dataset? Does the change of wind forcing explain the variability
of AABW export?
AR: Please, see the answer above for a similar question. We have included a sentence
talking about correlations between zonal wind stress and AABW export.
OSD
8, C731–C742, 2011
Interactive
Comment
(P1673/L21-P1674/L4) Does the model reproduce the absence of the AABW cold
pulse in the Northwestern Weddell sea in 2000?
AR: Yes, during the year 2000 potential temperature anomaly dropped to mean levels
(near zero) in the NW Weddell Sea; however the potential temperature anomaly remains negative in this region after 1997 in the model. We have added this sentence in
the m/s. Please see below.
"As well, potential temperature anomaly in the model (not shown) dropped to mean
levels (near zero) in the Northwestern Weddell Sea during 2000; however it remains
negative in this region after 1997."
Technical corrections: There are different expressions for deep ocean in the manuscript
(deep regime/ deep ocean regime/ open ocean regime/ oceanic regime), and they are
very confusing.
Full Screen / Esc
AR: We agree with the reviewer and changed all different expressions cited above to
"deep ocean" in the m/s.
Printer-friendly Version
[Fig.2 and its caption] The line for 28.27 of neutral density should be drawn.
Interactive Discussion
AR: The caption was wrong, the neutral density line corresponds to 28.27 instead of
28.26. We have corrected it.
[Reference] (P1679/L32) Validation of ***three*** global ocean models in the Weddell
C741
Discussion Paper
Sea
OSD
AR: Done.
8, C731–C742, 2011
Interactive comment on Ocean Sci. Discuss., 8, 1657, 2011.
Interactive
Comment
Full Screen / Esc
Printer-friendly Version
Interactive Discussion
Discussion Paper
C742